Is Inflation Coming Back? The PPI–CPI Gap Just Hit Its 97th Percentile
Research note. Computed from our own data pipeline — CPIAUCSL, PPIACO, CPILFESL, PCEPILFE, M2SL and FEDFUNDS — as of June 20, 2026. The exact methods are in How we checked it below; our editorial and AI standards are in How we use AI.
As of the May prints: CPI 4.2% and accelerating, PPI 13.1%, and the gap between them at the 97th percentile since 1948. Our read — real pressure, not yet entrenched. Core is still near 3%, which separates this (so far) from the 1970s. The swing factor is a Fed whose policy rate now sits below headline inflation.
Four months ago, US consumer inflation was 2.4% and falling. The last four prints: 2.4%, 3.3%, 3.8%, and now 4.2%. That alone would make “is inflation coming back?” a fair question.
But the consumer index is the slow gauge. Look upstream. Producer prices — the all-commodities pipeline that feeds everything else — have gone from 3.2% to 6.0% to 9.8% to 13.1% over the same four months. Quadrupled, in a season.
The distance between those two numbers is the story. Producer inflation is running 8.9 points hotter than consumer inflation. Since 1948 — 940 monthly readings — the gap has been wider in exactly 26 of them. That is the 97th percentile, and the company it keeps is not reassuring: 1951, 1973–74, 2008, and 2021–22.
Before getting excited: this PPI is supposed to swing
Before getting too excited about the gap, one thing matters: this is the all-commodities PPI, the rawest and most volatile cut of producer prices. It swings harder than CPI in both directions. That’s its nature, not a discovery, and comparing the two levels naively would be sloppy.
Which is why the right metric is the gap’s own history: how does today’s spread rank against every month of the same spread since 1948? Answer: 97th percentile. The volatility excuse is already priced into that ranking. When the noisy series detaches this far from the slow one, something real is moving through the pipeline.
Five episodes, three endings
Every prior visit to this territory resolved in one of three ways, and the difference between them is the difference between a commodity scare and a real inflation regime.
Ending one: the spike fades. 1951 (Korean War commodity hoarding) and 2008 (oil at $147). In both, the commodity surge that stretched the gap collapsed before it could fully transmit — CPI in 2008 peaked at just 5.3% before the financial crisis buried the question entirely. The gap closed from the PPI side. Consumers barely felt it.
Ending two: pass-through. 1973–74. Producer costs marched into consumer prices almost completely; CPI followed PPI up to 12.2% and stayed high for a decade because wages and services joined in. The gap closed from the CPI side — the bad ending.
Ending three: the hybrid. 2021–22. The gap blew out to a record 16.1 points. Consumer inflation followed producer prices up — CPI peaked above 8% — but then the commodity impulse faded and aggressive Fed tightening finished the job. Partial transmission, then resolution from both sides at once.
So which ending is this?
The case for “it fades”
Core inflation is not confirming. Core CPI sits at 2.8% and core PCE at 3.3% — drifting, not surging. So far the squeeze is being eaten upstream: producers and wholesalers absorbing costs rather than shelf tags repricing. The services-and-wages machinery that made 1973 permanent isn’t running yet, and that is the strongest argument this is a 1951/2008-shaped episode — violent upstream, muffled downstream.
The case for taking it seriously
The consumer index is already moving. Four consecutive accelerations to 4.2% is transmission happening, not transmission hypothetical — whatever is in the pipeline, some of it is reaching the register.
And the money side isn’t helping. Money supply growth has quietly reaccelerated to almost 5% a year — tinder that was absent in 2008 — while the federal funds rate sits parked near 3.6%. Against 4.2% headline, the real policy rate just went negative. The Fed that crushed the 2021 episode did it with a policy rate well above inflation. Officials are, for now, doing nothing.
A negative real rate, sustained while headline accelerates, is how the 1970s happened. Not in one quarter — through a year of “transitory” talk while the gap transmitted.
What it means for markets
The cruel part: the answer matters most for the things that are already stretched. We wrote last week that household equity allocation sits near records and margin leverage is growing at the 94th percentile. Positioning extremes are survivable while money stays easy. The historical cycle-ender is the tightening that inflation forces — and a Fed staring at 4%+ headline with a negative real rate is a Fed that will eventually be forced. That is what this gap is quietly working against: the easy-money assumption underneath every other extreme on our dashboard.
What we’re watching
This resolves print by print, and the next ones are dated: the June CPI lands July 14, PPI the morning after. We’ll take core CPI through 3.5% as transmission confirmed — that’s the red line — and PPI back under ~8% as the spike fading on schedule, which would make this a 2008-style footnote. Watch the yield curve too: the long end repricing before the Fed moves is how the bond market votes on entrenched-versus-transitory. Every chart in this essay is live and updates the day each series releases.
The uncomfortable part of our own read: the benign endings require the spike to fade soon, and policy is not doing much work yet. So — not the 1970s, core is too calm for that claim. But pressure at the 97th percentile, a consumer index already responding, and a central bank below headline is not a mix we’d lean against. Two of the three historical endings from here were benign. The one that wasn’t started exactly like this.
What this means for traders vs. long-term investors
Whichever of the three endings arrives, it lands differently depending on your horizon.
| If you’re a… | The read | What to watch |
|---|---|---|
| Trader weeks–months | A print-by-print trade. If transmission confirms, the inflation-sensitive side — energy and real assets, short duration, breakevens — is where it pays; if the spike fades, the disinflation trade wins. The asymmetry is that the Fed is already behind, with a policy rate below headline. | Core CPI through 3.5% (transmission confirmed) vs. PPI back under ~8% (the spike fading on schedule). The next CPI lands July 14, PPI the morning after. |
| Long-term investor years | The question is regime, not the next print. A sustained negative real policy rate argues for some inflation protection in the long-term mix — real assets, TIPS, less long-duration nominal debt — but a commodity-led spike that core hasn’t confirmed isn’t a reason to overhaul a plan. | Whether core and wages join the move — the 1973 path — and the real policy rate. That, not one hot PPI print, is what would justify a structural tilt. |
The read — A print-by-print trade. If transmission confirms, the inflation-sensitive side (energy, real assets, short duration, breakevens) pays; if the spike fades, the disinflation trade wins. The Fed is already behind, with a policy rate below headline.
What to watch — Core CPI through 3.5% (transmission confirmed) vs. PPI back under ~8%. Next CPI lands July 14, PPI the morning after.
The read — The question is regime, not the next print. A sustained negative real policy rate argues for some inflation protection (real assets, TIPS, less long-duration nominal debt) — but a commodity-led spike core hasn’t confirmed isn’t a reason to overhaul a plan.
What to watch — Whether core and wages join the move — the 1973 path — and the real policy rate. That, not one hot PPI print, justifies a structural tilt.
The trader is positioning for the next print; the investor is deciding whether to tilt a whole portfolio. Both turn on the same hinge — whether core confirms what commodities are already shouting.
How we checked it
None of these numbers are estimates — each comes straight from the official price series. Here’s how, in plain terms:
- Year-over-year rates use each index’s own official 12-month change.
- The “gap” is simply producer inflation minus consumer inflation, month by month.
- The 97th-percentile claim means today’s gap is wider than 97% of all monthly readings since 1948.
- The historical episodes are the actual months the gap was this wide before (1951, 1973–74, 2008, 2021–22); what each did to consumer prices afterward is reported, not guessed.
Frequently asked questions
What is the difference between PPI and CPI?
CPI measures prices consumers pay at the register. PPI measures prices producers receive — here the all-commodities index, the rawest upstream read. PPI captures cost pressure in the pipeline before it reaches consumers, which is why large PPI moves are watched as a leading signal.
Does PPI lead CPI?
Loosely, and with slippage. Producer costs either get passed to consumers (CPI follows PPI up, as in 1973–74), get absorbed in profit margins, or the commodity spike that drove PPI fades before transmission completes (1951, 2008). The gap between the two is the tell worth tracking, not either series alone.
How unusual is the current PPI–CPI gap?
As of the May 2026 prints, PPI is rising 13.1% year-over-year against CPI at 4.2% — a gap of 8.9 points, the 97th percentile of all monthly readings since 1948. Only 26 of 940 months were wider, clustered in 1951, 1973–74, 2008, and 2021–22.
Is core inflation also rising?
Much more slowly. Core CPI is 2.8% and core PCE 3.3% — drifting up, not surging. The current episode is so far commodity-led rather than broad-based, which matters because the 1973-style bad outcome ran through services and wages, not just goods.
What is the Fed doing about it?
As of publication, standing still. The federal funds rate has been flat near 3.6% for months, which means the policy rate is now below headline CPI — a negative real rate against headline inflation. Historically that mix, sustained, is how moderate inflation becomes entrenched inflation.
Our copies (browsable, with CSV download): CPI, PPI, the full indicator library.
Spot an error? Email info@thetrading.tools — we correct on the page and bump the modified date. Educational content, not financial advice.